GENE-ENVIRONMENT INTERACTION EFFECTS ON EARLY BRAIN DEVELOPMENT AND GENETIC BASIS OF BRAIN ORGANISATION

Abstract

Firstly, we developed a set-based mixed effect model for gene-environment interaction (GxE) which we named MixGE. There has been no mixed effect GxE model available prior to MixGE, and MixGE is both statistically powerful and computationally efficient. MixGE revealed the interactive effect of ABCA7 and cardiovascular risk on local volume change of right superior parietal cortex. We also tested and supported our hypothesis that FKBP5 genotype moderates the association between antenatal maternal depressive symptoms and in utero hippocampal development. Secondly, we formulated a procedure to parcellate the left cerebral cortex based on gene expression profile, which has never been achieved before. We showed that gene expression parcellation better reveals the molecular support for structural and functional connectivity than parcellations based on other data modalities. We also showed that neurotransmitter pathways are significantly overrepresented by genes that drive molecular organisation, and we produced gene expression maps of various neurotransmitter receptors.